Urinary cortisol in the assessment of pituitary-adrenal function: utility of 24-hour and spot determinations

To more conveniently assess dynamic changes in the biologically active fraction of cortisol, we measured cortisol in 1-h urine samples obtained from 0700-0800 and from 2200-2300 h. In 20 normal subjects, morning 1-h urinary cortisol levels were 78 +/- 36 ng/mg creatinine (mean +/- SD), whereas levels from 2200-2300 h were 22 +/- 12 ng/mg creatinine, demonstrating diurnal variability. In 14 patients with Cushing's syndrome, mean morning urinary cortisol was elevated (207 +/- 176 ng/mg creatinine), but there was overlap with values in normal subjects. In contrast, evening values in Cushing's syndrome (248 +/- 208 ng/mg creatinine) were elevated in each patient; there was no diurnal variation and no overlap with normal subjects. Similarly, the morning urinary cortisol response to dexamethasone (1 mg, orally, at 2300 h) clearly separated normal subjects from those with Cushing's syndrome (5 +/- 6 vs. 169 +/- 149 ng/mg creatinine, respectively). In 10 patients with secondary hypoadrenalism, urinary cortisol levels were less than 2 ng/mg creatinine in both morning and evening 1-h samples. Thus, the determination of cortisol in 1-h samples is a practical and simple method of assessing cortisol secretion and allows multiple sampling without hospitalization. It is effective in assessing dynamic cortisol responses, such as diurnal variation and responsiveness to suppression, and it is an effective screening test for Cushing's syndrome and hypoadrenalism.